JPS6343461A - Detection system for response of called person - Google Patents

Abstract

PURPOSE:To detect an response from a called person by detecting proper frequenct component contained in various signal tones arriving from a telephone circuit separately in a terminal equipment and watching time of continuation. CONSTITUTION:In this telephone circuit, even when a called person responds, metering signals such as polarity reversion pulse etc. do not come in. Detection of response of the called person is made by detecting signals arrived from the telephone circuit for each frequency component by an FDT 8, and observing whether the detection output Df continues for a specified time or not by a CNT 7. That is, the CNT 7 judges various signal tones and voice of the called person, and detects response of the called person by judgement of voice of the called person, and controls a CPS 9 according to it and receives coins.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a method for automatically detecting a response from the other party in a terminal device connected to a telephone line in which a signal indicating a response from the other party does not arrive. It is something.

[Conventional technology]

Dialing is generally done through one congregation telephone line;
When the other party responds, a signal indicating the other party's response, such as a polarity reversal pulse, is sent from the exchange, and a signal indicating the other party's response is received from the telephone port n. It is said that L1 gold is performed according to the (ff number) of pulses, etc.

On the other hand, in some foreign countries, even on telephone lines connected to public telephones, there are cases where signals such as polarity reversal pulses do not arrive at all when the other party responds, and in this case, conventional means are used. In some cases, billing becomes impossible, so special means are used to detect the other party's response.

That is, the calling party's handset is left inactive,
A means for detecting the voice of the other party using an amplified detector @ and determining whether the other party is responding, or detecting the ringing tone in the same way, and determining that the other party is responding according to the elapse of a certain period of time from the end of the detection. A method or a combination of both are generally used.

[Problem that the invention seeks to solve]

However, in addition to the voice of the other party and the ringing tone, the signals arriving from the telephone line include intercept sounds such as dial tone, busy tone, and talkie guidance. Misjudgment may result in unnecessary charges being made, causing unexpected damage to users.

[Means for solving problems]

In order to solve the above-mentioned problem, the present invention is constructed by the following means.

That is, in the above-mentioned terminal equipment, each unique frequency component included in various signal tones arriving from the telephone line is individually detected, the duration of each detection force is monitored, and the specified period of the specific station wave number component is detected. Continuing, it is determined that the signal sound corresponds to this condition, and when none of the frequency components can be determined to be various signal sounds, it is determined that it is the voice of the other party, and based on this determination, the response of the other party is detected.

[Function] Therefore, depending on the presence or absence of the same wave number component unique to each ring signal sound and the duration of the detected frequency component, it is determined which signal sound has arrived. Since the duration of the specific station wave number component is discontinuous and cannot be determined to be any signal, it is determined that the signal is the voice of the other party, and thereby the response of the other party can be detected.

〔Example〕

Hereinafter, the present invention will be explained in detail with reference to figures showing examples.

FIG. 2 is a block diagram of a public telephone set, in which line terminals L1. L.

In contrast, hook switch H81, receiving circuit (hereinafter referred to as RE
C) 1, Teriode bridge (hereinafter referred to as DB) 2,
A communication loop is formed via a power supply circuit (hereinafter referred to as PS) 3 and a forced disconnection circuit (hereinafter referred to as CB) 4, and a call-free circuit (hereinafter referred to as T
KO) 5 is connected, this includes the handset 6.
A transmitter T and a receiver R are connected.

Processors such as microprocessors (hereinafter referred to as CP)
control unit (hereinafter referred to as CNT) γ consisting of U) and memory, etc.
A frequency detection circuit (hereinafter referred to as FDT) 8, a coin processing unit (hereinafter referred to as CP
S) 9, dial key (hereinafter referred to as DK) 10, hook switch Ha2, etc. makes control decisions according to each output, CPS 9, display unit (hereinafter referred to as DP) 12, audio composite 1 IC or less, MF) signal 1F signal generator (hereinafter referred to as MFG) 13, TKC5, and CB
4, etc.

In addition, MFG 13 outputs DKlo and CNT1.
MF responds to the control of DKlo and follows the operation of DKlo only when the permission signal OK is given from CNT7.
When a double signal is generated, a control signal cst'' is given to TKC5, and the MF double signal dial signal is sent via TKC5, and monitoring information is sent to a maintenance center etc., rarG13 follows the control of CNT7. After making a dial call to a specific destination such as a ski center in the same manner as described above, monitoring information is transmitted via the same route using the W signal.

On the other hand, PS3 has hook switch I using off-hook.
When ts1 is turned on, a DC loop is closed via TKC5, and a loop current with a constant polarity passes through DB2, which charges the capacitor and uses the terminal voltage of the capacitor as a power source. It is intended to be supplied to each department.

Therefore, if off-hook and coin insertion are performed, the DC loop is closed and CNTγ sends out a permission signal OK in response to the output of CPS9, so DKlo
It is possible to make a dial call, and depending on the other party's response, the call can be made freely.

However, even if the other party responds, the f! On the telephone line, billing signals such as pulse pulses do not arrive, and in order to detect the other party's response, the incoming signal from the telephone line, which is given via the REC1 using a converter, etc., is analyzed for each same wave number component by the FDT8. The CNT 7 determines whether this detection output Df continues for a predetermined period of time, and based on this result, the CNT 7 judges various signal tones and the voice of the other party. Detects the other party's response by judging the other party's voice, controls cps5 accordingly, stores the coin, eliminates the mute signal MUT for TKC5, and releases the inoperable state of the transmitter T. On the other hand, if there are not enough coins waiting to be inserted, CH2 is controlled to open the DC loop for a certain period of time, restore the exchange, and forcibly disconnect the call.

Depending on the DP12, the CNT7 sequentially displays the coin insertion and accumulation status, while the CNT7 has a self-diagnosis function and performs self-diagnosis according to the output of the hook switch H82 in response to an off-hook condition. It detects jammed coins, disconnection of the handset 6 due to theft, etc., and accordingly displays on the DP 12 that p-mail cannot be used, and also sends the above-mentioned monitoring information.

Furthermore, in F'DT8, the lowest detection level of each frequency component is freely variable according to control signal CL, and CNT7
controls the FDT8 according to conditions, and successively sets the lowest detection level to a desired value.

In addition, a busy tone generator (hereinafter referred to as "BTG") 14 that locally generates a busy tone is provided, and BTGf 4 sends out a local busy tone according to the control of the CNT 7 and converts the transformer T. The signal is sent to the receiver R of the handset 6 via the receiver R to notify the user that the call is not possible.

FIG. 3 is a block diagram of the FDT 8, including a preamplifier (hereinafter referred to as HA) 21 and a fixed attenuator (hereinafter referred to as PAD).
22. Variable attenuator (hereinafter referred to as ATT) 23. Limiting amplifier (
Hereinafter, LA) 24, PAD25, and selection amplifiers (hereinafter referred to as 5A) 26+ to 2, each having an individual selection frequency.
66, and amplified detectors (hereinafter referred to as AD) 27s to 2711 with built-in Schmitt trigger circuits constitute the main system, and these are integrated circuits,
PAD22 and ATT23 can be freely set with insertion attenuation circuits (hereinafter referred to as ASW) 281 to 284 such as analog switches! +, ASW28t~28
4 is off, insertion attenuation flOdB, ASW2
When only 8+ is on, the same attenuation amount is 5 dR1ASW 28
When only z is on, the same attenuation is 10 dB, ASW 28 s
When only ASW284 is on, the attenuation is 20 dB, and when only ASW284 is on, the attenuation is 30 dB.
determining combinations of the control signals CLs to CL4 from the NT7 in correspondence with on/off of the ASWs 28t to 284 corresponding to the lowest detection level (hereinafter referred to as LDL);
It is possible to set the LDL for the @frequency component of FDT8.

Also, when the insertion attenuation of PAD22 and ATT23 is OdB, L[)L is -554 at the input of HA21.
BV Terror, D, PA [) 25 O and ATT 23
From each insertion attenuation combination K, -55 dBV ~ -2
In 5dB steps between 0dBVO! 7LD
L is determined, and the following code is assigned to each LDL for control by CNT7.

On the other hand, various signal tones coming from the telephone line, in this case,
Each frequency component of the next light is included as an individual component, and the optimal LDL for each detection has the relationship shown in the next light.

Table 2 - 8 sounds (hereinafter referred to as D), ringtone (hereinafter referred to as RB)
T), busy tone (hereinafter referred to as BT), and intercept sound including chime sound (hereinafter referred to as INT), each of which continues for 1'lU for a predetermined period of time in which the visible same wave number component appears, and RBT, B
In contrast, in the case of T, the appearance of each frequency component is unspecified, and each judgment is made according to these conditions, whereas in the case of voice, the occurrence situation of each frequency component is unspecified.

Therefore, in FIG. 3, the signal arriving from the telephone line via RECI is amplified by HA21, and P
After giving A[)22 and ATT23 a fluid reduction corresponding to LDL K, the imaging width is limited at LA24, and it is given to each 5A26t to 26s via PAD25, thereby creating a valley of 350Hz = 1912Hz.
! Extract the 3 wave number components and calculate the trough extraction force from AD27t to 27
After detecting a predetermined level or higher by g and amplifying it, it is sent to CNT7 as 8-bit detection power D11 to DAB, each of which shows a binary value, and a control signal C from CNTγ.
After receiving the LDL settings according to LI~CL4,
The CNT 7 shown in FIG. 2 performs the above-mentioned judgment and detection depending on the conditions of the detection forces Df1 to D□ and their joint time.

FIG. 1 is a comprehensive flowchart of the judgment and control situation by the CPU in the CNTr, which is executed while executing instructions in the memory and accessing necessary data from the memory.

In other words, when the control signal CL is sent out, LDL=7J.
101, and set the detection function "Df import" 1.
02, check "Df large input shift" 103, and if this is Y (YES), control signal C
Due to LO change, D r LDL =, 4 J 104
,l! : After raising LDL, r LDL≧7
? LD of FDT 8 via N (No) of J 111
In order to estimate the time required to stabilize the operating condition in response to the L change, configure it in the CPU, perform (r30ms timer set) 112, and count the number of occurrences of each detected power Dfi to Df8. After performing the "Frequency specific counter clear" 113 provided in step 102, "Df import" 114 is performed again as in step 102, and as in step 103, "Df large input? J 12
Check 1, if this is Y″″C, step 1
12 r30ms timer 〇 clear” 122,
In step 113, the "frequency-specific counter count-up" 123 is carried out in correspondence with the detection power Dll to DIB. After that, the detection power Dfl to DAB is performed.
5X
Since lO = 50ml is assumed to be one station, r50m
s cycle/end? ” 124 is N, steps 114 and subsequent steps are repeated.

Takata, when step 121 is N, is the r30ms timer Φ corresponding to step 112 being set? J131
Assuming OY, r30ms timer -IJ132
Perform subtraction and the content becomes zero, Qr3
0ms timer eEND? J 133 becomes Y, rLDL= LDL + lJ 134 j7
The LDL is changed by t-1 stages to lower the LDL, and then steps Ill and subsequent steps are repeated. When the LDL reaches 7 and step 111 becomes Y, the process returns to step 101 and the subsequent steps are repeated.

On the other hand, as step 124 becomes Y, based on the count value for each frequency component in step 123, the frequencies 1150HS and 1566Hzs 1 in Table 2 are determined.
800Hz, 1912I (z is 10/1 as the detection frequency
0 or not? INT=10/10? J 141 judgment, same wave number 420Hz, 490Hz 1620
r420Hz-10 depending on whether Hz is 10/10 or not
/10? J 142, r 490Hz = 10/
10↑"14λr620Hz=10/10]" 144 are performed sequentially, and if these are Y, it means that each frequency component has been stable for a time of 50 m5 or more, and according to Y in step 141 rINT processing" 151, and depending on Y in step 142 or 143, "RBTBT processing" 151.
Go to process 52, and in step 144, "BT process" 15
Step 3 individually.

For the above, if steps 141 to 144 are all N, the same as steps 102 and 103 <"Import with D" 1
61 and rDf input/reed? The effective data length r configured in the CPU in order to make the judgment in 162 and monitor the duration of the detection force that should be judged as valid. The gap is “EFC timer ≧200m5?
Step 101 and subsequent steps are repeated until N is selected in J 164, and if step 164 becomes Y, the process moves to "audio processing" 171.

Therefore, 4! of the signal coming from the telephone line! The r frequency components are individually detected, and the outlines of various signal tones are determined according to their continuation status.If they are not distinguished from any of the signal tones, the outline is determined as the other party's voice, More detailed judgments will be made based on these summary judgments.

FIG. 4 shows the lower routine of ``INT processing'' 151, and the ``01 human power
END? If J 201 is Y, step 163 r EFC timer≧200m5? As J 201 hits Y, the duration of detection force Df7 is r 200
m5≦1.8KHz≦400m5? J 203 is checked, and if this is N, the process moves to "audio processing" 211, whereas if step 203 is Y, INT
Therefore, the INT logging process from step 221 onward is carried out.

That is, it executes "209 Timer Sera) J221" configured in the CPU, continues sending the mute signal MUT to the TKC5, maintains the state of "Telephone operation continues to stop" 222, and sends display data to the DP12. Send ``Please hang 1 handset'' Display J
After performing 231e, in step 221 "20s timer -END? J In response to 232 becoming Y,
Controls BTG 14 and sends “local BT transmission” 24 to handset R
1, and the output of hook switch H82 is #) [
When the on-hook T J 242 becomes Y, the series of controls ends.

Therefore, IDoHz% 1566Hz% 180
Each frequency component of 0Hzs1912Hz has a repeating time of 200m5 or more, and the frequency component of 1800Hz is 200m5 to 400m5 (if divided by D duration,
A determination is made that it is INT, and the display in step 231 is 2.
After a period of 0s, a local BT is sent to notify the user that the call is not possible.

FIG. 5 shows the lower routine of "rRBT processing" 152.
END? If J 301 is Y, step 16
3 r gFc timer ≧2.58? ” 302, and assuming this is N, the frequency of occurrence of the detection force D14 is r 620Hz = O/10? J303
If this is N, step 302
Similarly, rgvc timer ≧200m5? J304
is checked, and returns to step 101 during N. However, when step 304 becomes Y, for example, a 6s r RB configured in the CPU is configured in the CPU to monitor the interruption time during the RBT interruption time. T break timer set 305 starts this, and then by sending the control signal CL, 5 r LDL =
Continuously configure the settings for 7 pieces 311, and determine the detection force "Df large input? J312, and if this is N, enter r RBT break timer - END? J3 in step 305.
13, and step 311 according to the value of N.
While repeating the following steps, as soon as step 312 becomes Y, set ``r LDL = RBT detection level'' 312 in the same way as step 311 based on Table 2, and then set ``r LDL = RBT detection level'' 312 as in step 112, and then set 315, and the same as step 312.
Df large input afi? Check J 321 and specify the duration of data that is not noise and is recognized as valid (according to J 322).
If this is N, use 9r1 in step 322.
Step 3 via "50ms timer clear" 323
Repeat steps 21 onwards.

Further, while step 321 is N, step 315 corresponds to r 30mg tie-r--gND? If j324 becomes Y, r LDL = LDL + 1''325
The control signal CL is updated to lower the LDL, and steps 315 and subsequent steps are repeated.

For the above, steps 321 and 322 are both Y
In this case, since the first detected duration during the RBT intermittent time is unspecified, in order to confirm the interruption time, first the first interruption time is stored in rRBT・first time・memory storage” 331. And this is r RBT・1st@time≧2.5 a
? J 332 and the same as step 301 <"Df large input ID? j As 333 becomes Y, the duration following the first new period is r RBT-
@i Duration time/memory storage" 334, same as step 305 < r RBT break timer/
After performing "set" 335, the contents of step 334 are rRBT-first duration ≧700m5? J34
1, and if the result is N, r.
343.

In addition, Y'! of step 313! Otherwise, the process moves to step 342 also by N in step 332.

If step 341 is Y, the same as step 311 (rLDL=7J'351 is set, and step 3
12. Same as 313 <“Df large input a#)T” 352,
rRBT break timer・E:ND? J353
and in response to Y in step 352 are the same as steps 314 to 325 (r I, DL = R
BT detection level" 354~rLDt=LDL+1J36
5, and the second and subsequent "RBT/nith
1 interruption time/memory storage" 371, and each interruption time of the nth and n+1th time is r RBT・W, the n+1th interruption time in the n interruption time? , J 372, this is Y
If so, as in step 333, “Df large input END?
” 373 and step 3 according to Y of this.
34 Similarly, the second and subsequent r RIIT・n+1
Duration/memory storage” 374, step 3
Same as 72, each duration of the nth and n110th r
RBT-n+1st duration in nth duration? J37
5 Compare again and if this is also Y, step 335
Same as (r RBT break timer set" 381
After performing this, steps 351 and subsequent steps are repeated.

Note that even if N at step 353 is determined, step 3 is immediately executed.
On the other hand, depending on step 3530Y1 or N in steps 372 and 375, the process returns to step 342.

By the way, 420Hz? Take 490Hz tD
MJ wave number component continues over 150m5, continues, 2.5+
If the above interruption time and the duration time are 700m5 or more, by comparing the latest one and the immediately preceding one, whether or not each interruption time is the same or not. This comparison is repeated, and when the RBT stops due to the other party's response, the process proceeds to the "voice detection process" described later. Transition.

FIG. 6 is a lower routine of "rBT processing" 153, and when the detecting forces Df1 to Df6 disappear, "Df large input END? If J401 becomes Y, rgrc timer ≥ 2.5s? J402, r EFC timer≧
700m5? J 403, r grc timer ≧20
0m5? Check J404 as a similar fire, step 40
2. 403 is N1 Step 404 While N returns to step 101, depending on Y in step 404, the r 700rns timer set in the CPU 41
1, and assuming that the detection forces Df1 to Df6 are N in "Df large input? J412," r700ms timer write END?
3 is checked, and while this is N, steps 412 and subsequent steps are repeated, and as step 412 becomes Y,
Since the first detected duration during the BT intermittent time is unspecified, the first intermittent time is stored in "BT/first intermittent time/memory storage" 414.

Next, as in step 401, "Df large input F,
ND? In response to Y in J 421, the duration following the first predetermined period is stored in r'BT/first duration/memory storage' 422, and similarly to steps 412 and 414, 'D
f large input foot? , according to Y of J431, set the second interruption time r BT"@22 interruption/memory storage" 43
2, and then calculate the first and second interruption times r.
BT-1st break time * 2nd break time? Compare using J433.

If step 433 is Y, “Df person car END?J
In response to Y in step 441, the second duration is stored in "rBT - Second duration - memory storage" 442, and in the same manner as in step 433, rBT - Second duration in first duration? It is dangerous to perform a comparison using J443, and if this is also Y,
The process moves to a predetermined step during "second DT processing" to be described later, and local BT transmission to the handset R is performed.

Also, if step 403 or 413 is Y, rR
After performing BT detection flag/cera) J451, the process moves to step 305, and while step 404 is N, the process moves to step 101, while in response to Y in step 402 or N in steps 433 and 443, the process described below is executed. The process moves to “audio detection processing” 452.

Therefore, the frequency component of 620Hz is 200m5~
If it continues for 700 m5, each intermittent time is compared twice, and if they match, the on and off times are each 0.5
s'! Intermittent BT is determined based on a fixed cycle of 0.25 s each on and off.If it is 700m5 or more, it will be processed, and if it is 2.58 or more, it will be processed. Move to "Processing".

FIG. 7 shows the lower routine of “audio processing” 171,
r zFc timer ≧2.5 s? J501
is N while r gFc timer ≧700m5? J50
．． 2 is checked, and if it is 4N, the "voice detection process" from step 511 onwards is performed.

That is, "73 timer set" 511 in the CPU
After starting this operation and performing ``cancellation of transmitter operation stop'' 512 to stop sending out the mute signal MUT, perform the [Tone processing 31J 513 in FIG. Timer/END
? J514 is determined, and while this is N, "On hook?" is checked in response to the output of the hook switch H82. If this is also N, steps 512 and subsequent steps are repeated.

If the result in step 514 is Y, it is determined that a call with the other party has started, the CPS 9 is controlled to perform "coin storage" 521, and an OK signal is sent to allow the request for data service by sending an MF signal. After the "data service request permission" 522 is executed, the "20th
T processing" 523 is performed, and as in step 515, [
On hook? J 524 is checked and a series of controls is completed according to Y of this.

On the other hand, depending on Y in step 502, r RBT detection flag Sera) J 531 is performed, and
The same flag is also set in "BT processing" 153, and when step 515 is Y, the r RB
T detection flag e set? J541 is checked, and if it is Y, the process moves to step 521, whereas if step 541 is N, CPS9 is checked to avoid erroneous storage.
After performing "coin return" 542, the series of controls ends.

Therefore, "voice detection processing" is performed through the conditions shown in Figure 1 or 4, or each of the conditions in Figure 5 or 6, and as a general rule, the voice is detected by "tone processing" during the 78 period following the stop of RBT. Reconfirmation is performed including the change of LI)L, and if it is not determined that it is a signal tone during this time, the other party's response is detected by the voice of the call, the coin is stored, and on-hook is made between 7a. However, if the RBT detection flag is not set, the coin will be returned.
Unexpected damage to users is avoided.

FIG. 8 shows the lower routine of "second DT processing" 523. In some foreign countries, the DT is sent again when the other party goes off-hook. In this case, In particular, these processes are performed for the purpose of preventing disconnection, since dialing can be made without repeating on-hook and off-hook operations.

That is, based on Table 2, r LDL = DT detection level 601 is set for sending out the control signal CL, and in consideration of the use of a hand-held MF signal dial transmitter, the detection power Df1 ~ Df3 Yoshi 350.400
．． 440, 480, 500Hz, etc., and extract
T large input a, ri? In response to Y of J 602, “D
T input ≧20m5? After determining J603 and setting the r loomg timer Φ J611 for a time that does not affect the call, "stopping the transmitter operation" 612 is performed by sending a mute signal MUT, and further step 602 Same as "DT large input available? J61
2, and on the assumption that this is Y, in step 611 "1100rn timer [END? J 614 is N"
During this time, steps 613 and subsequent steps are repeated.

Further, if step 613 is N, mute signal MU
When the transmission of T is stopped, "Telephone operation stop cancellation J621.
Then, in correspondence with step 611, "100ma timer clear" 622 is performed, and then steps 602 and subsequent steps are repeated.

On the other hand, as step 614 becomes Y, DP1
Send display data to the receiver RK, perform a "Please hang up the handset" display 631, control the BTG 14, and then control the BTG 14. On the hook?”
When 633 becomes Y, the series of controls ends.

For this reason, M2D that arrives in response to the on-hook of the other party
Detected in response to a continuation of 20 m8 or more of the MF multiplied signal via the transmitter T from a hand-held MP signal dial transmitter, and by making the transmitter T inoperable for 100 m5, Redialing is blocked and local B
By transmitting TO, the user is notified of the inability to make subsequent calls.

7'j XL, dead period with transmitter is 100nxs
During this period, even if the transmitter T is inoperative, there will be no particular hindrance to the telephone conversation.

Therefore, by setting the LDL for FDT8K, detecting the 1-5r station wave number component unique to each signal tone based on this, and monitoring the duration of each frequency component, it is possible to determine the signal tone corresponding to each of these components. is determined, and if none of the frequency components can be determined to be a signal tone,
It is determined that it is the other party's voice, and the response from the other party is thereby reliably detected.In particular, is it possible to charge billing and call even if a signal indicating the other party's response does not arrive? Can be done accurately.

However, the frequency at which the frequency component is detected can be determined according to each signal tone and the transmission frequency band of the telephone line, etc., and each timer for monitoring the duration can be set according to the intermittent status of each signal tone and the detection required. You can set it according to the time.

In addition, the configurations in Figures 2, w, and 3 can be selected arbitrarily depending on the situation, and in Figure 1 and Figures 4 to 8, the steps are replaced according to the judgment conditions. Various modifications are possible, such as replacing the steps with other equivalent steps or omitting unnecessary steps.

〔Effect of the invention〕

As is clear from the above explanation, according to the present invention, by detecting frequency components and monitoring the duration of a specific station wave number component, the other party's Since a response can be reliably detected, a remarkable effect can be obtained in various terminal devices such as public telephones and facsimile machines that perform billing using coins, toll cards, etc.

[Brief explanation of the drawing]

The figures show an embodiment of the present invention, in which Figure 1 is a comprehensive flowchart of judgment and control situations, Figure 2 is a block diagram of a public telephone, Figure 3 is a block diagram of FDT, and Figures 4 to 8. 2 is a flowchart showing the lower routine of FIG. 1. FIG. 1 work...REC (receiving circuit), 7...CNT (
control unit), 8...FDT (1!I wave number detection circuit),
22...PAD (fixed attenuator), 23...A
TT (variable attenuator), 281-26...SA
(selective amplifier), 271-278 ・AD (amplification detector), 2L-284 ・・Fist-ASW (switch circuit), L hL 1 6 @ @ @ line terminal, D
fr DiL~Dfg... detection force, CL, CL
1~CL4 ・Reference・Control signal.

Claims (1)

[Claims] In a terminal device connected to a telephone line where no signal indicating a response from the other party arrives, each unique frequency component included in the various signal tones arriving from the telephone line is individually detected, and the duration of each detection force is determined. It is determined that the signal tone corresponds to this condition by the continuation of a specific frequency component for a predetermined period of time, and when none of the frequency components can be determined to be the various signal tones, it is determined to be the voice of the other party, and based on this determination, the response of the other party is determined. A partner response detection method characterized by detection.